Smart Sales Helper

• 

  Smart Sales Helper

• 

  Initiating bot & understanding users

• 

  Sales data analysis

Inspiration

We have the opportunity to innovate and think creatively at this event. It provides our team with a chance to enhance and learn new skills. The workshops offered are engaging and help us improve our technical knowledge. Additionally, we get to work on solving real-world problems.

What it does

The bot is designed to assist the TikTok sales team by analyzing sales data and providing useful summaries and predictions. Its main functionalities include data analysis and sales forecasting, which can help the sales team make informed decisions.

How we built it

We used Python to create a Telegram bot that interacts with users by leveraging the Telegram API. This bot captures user information and performs data processing and machine learning tasks, storing results in a database. We began by setting up the environment necessary for our project. This involved installing and importing various Python libraries. We initialized the Telegram bot using a bot token, which is stored as an environment variable for security purposes. The telebot library handles the connection to the Telegram API, allowing the bot to send and receive messages. We separated the machine learning tasks from the bot's interaction logic for better organization and modularity. Finally, we combined the bot's interaction logic with the machine learning model. When a user requests a prediction (using keywords like "prediction" or "forecast"), the bot processes the request.

Challenges we ran into

Initially we planned to used larkbot, but was faced issues with the setting up, but with the lack of time, we decided to switched to telegram bot instead. We challenged ourselves and tried to conduct machine learning which we do not have much experience of, and hence we used quite a large number of time to research on how to predict sales data

Accomplishments that we're proud of

We were proud that our telegram bot managed to work, and could reply to our messages, and send images of our predictions to the telegram bot

What we learned

We have learnt that time management is very important, as time was tight. We learnt how to create a machine learning model and the importance of data cleaning

What's next for Smart Sales Helper

We hope to further improve our telegram bot, to incorporate telegram groups chats amongs teams, so that we can gather feedback by managers and give even more accurate and effective sales improvement to agents.

Built With

• python
• telegram